Agentic litigation-investigation platform. Matter workspace, evidence mapping, cross-evidence contradiction detection, trial-bundle compilation, precedent bank, voice dictation, court-form pack — all built around a per-matter reasoning engine with inspectable audit trails. Live in production.
Pillar 01Live in productionUK (E&W) + PLPer-matter isolation
What it does
— 01
Pulls pleadings, witness statements, disclosure, correspondence, and exhibits into one matter workspace. Surfaces contradictions between witness accounts, timeline gaps, and documents that rebut each other. The pre-trial analysis a junior would spend weeks on, done continuously in the background.
— 02
Generates court-ready trial bundles from the matter workspace. Pagination, index, cross-references, exhibits. Handles UK CPR-compliant format and Polish KPC format. What used to take a paralegal a week takes the system minutes — with every page traceable to its source document.
— 03
Per-firm source-of-truth memory of house positions, negotiated carve-outs, and prior-matter outcomes. Retrieval grounded in the firm's own playbook, not a generic web corpus. Builds a compounding asset with every matter you run.
— 04
Solicitor dictates analysis or instructions at the point of thought — the system transcribes, attributes to the matter, and routes to the right workflow. Three engines supported (browser Web Speech API, Whisper, provider-redundant fallback) so the feature keeps working when any one upstream is degraded.
— 05
20 Polish civil-procedure court forms bundled as structured templates with quick-fill from matter data. UK equivalents (N-series, claim forms, application notices) on the roadmap. Filing becomes data entry once, not re-keying into PDFs every time.
— 06
Solicitor-client privilege preserved across agent interactions. No cross-matter bleed. Every output carries a full trace — which agent produced it, which sources it consulted, which verification rules ran — defensible for professional-conduct reviews and insurer queries.
Who it’s for
Litigation-heavy firms — UK and PL
The firms where evidence triage and bundle preparation eat the most partner and paralegal time. Legal AI handles the investigation stack end-to-end: evidence map, contradiction detection, precedent retrieval, bundle compilation. What used to need a dedicated paralegal becomes one seat + platform. Solicitors keep judgment; the platform handles the paperwork.
Sole practitioners and small firms
The matter-workspace tier. Court-form pack, voice dictation, basic precedent memory, calendar export, encrypted backup. Replaces the patchwork of Word templates, PDFs, and Outlook that the sole practitioner operates today — without the enterprise price tag of Harvey or Lexis+.
How it’s built
Not a single prompt or a chatbot. A composed pipeline of specialised agents, each with a defined role, a documented toolset, and a declared failure mode.
agent/ingest-matterParses the incoming matter pack — pleadings, disclosure, witness statements, correspondence, exhibits. Identifies matter type, parties, jurisdiction, governing law. Produces a structured matter record the rest of the pipeline works against.
agent/map-evidenceBuilds the evidence map — cross-references between documents, timeline reconstruction, attribution of claims to sources. Produces the spine of the trial bundle without a paralegal keeping it up to date by hand.
agent/cross-evidence-qaIdentifies contradictions between witness accounts, timeline gaps, documents that rebut each other, and claims unsupported by the evidence base. Flags for partner attention before cross-examination, not during.
agent/retrieve-precedentQueries the per-firm source of truth — prior matters for this client, firm house positions, negotiated carve-outs. Reasoning grounded in the firm's playbook, not a web corpus.
agent/compile-bundleGenerates the trial bundle in court-ready format. UK CPR pagination and index rules. Polish KPC format where applicable. Every page traceable to source.
agent/qa-complianceCompliance gate. Pre-filing check against CPR, KPC, SRA, adwokacka/radcowska tajemnica zawodowa, and RODO/UK GDPR rules. Rejects or flags anything below confidence threshold — professional conduct, not commercial QA.
Status
Status
Live in production. Deployed against live matters since early 2026. First external solicitors onboarding through 2026.
Jurisdictions
UK (England & Wales — civil, commercial, immigration, family). Poland (KPC — sądy powszechne, 20 court forms bundled). EU English-language work supported.
Commercial model
Two tiers. Tier A — full litigation-investigation platform. Tier B — matter workspace + court-form pack + basic precedent memory. Per-firm licence, not per-seat. Founding-member engagement open for the first 100 firms per market.
Compliance
UK GDPR + EU GDPR (RODO). SRA-aligned audit trail. Adwokacka / radcowska tajemnica zawodowa preserved. Per-matter isolation. Full DPA on request.
Engage
enterprise@blackflake.com — include practice area, fee-earner count, and primary use case.
How we engage
Every deployment is a bespoke engagement. We don't sell seats or generic SaaS. We build agents against your workflows, your matter taxonomy, your regulatory context — then we operate the system with you under a monthly retainer. Your data, configurations, and outputs are yours and portable. The Blackflake engine is developed by us and runs under your retainer.
— 01
Scoped call to map your workflow, constraints, and success metrics. No charge. You leave with an architecture note and engagement estimate.
— 02
Fixed-price deployment on one narrow workflow. Real output on real work in weeks. Go / no-go before the full build.
— 03
Agent suite, workflows, sources of truth, QA gates, operator surface. Deployed in your tenant. Auditable.
— 04
Monthly retainer: new regulations ingested, agents tuned on live work, new workflows added. We develop the engine; your data, configurations, and outputs stay yours — always portable.